Rémi Pautrat
Senior Research Scientist at Microsoft
Zurich, Zurich, Switzerland
Summary
🤩
Rockstar🎓
Top SchoolRémi Pautrat is a Senior Research Scientist at Microsoft’s Mixed Reality & AI Lab in Zurich with eight years of experience in computer vision, spatial AI and 3D scene understanding. He blends academic rigor from a PhD at ETH Zurich with applied R&D on visual localization, mapping and multimodal neural training, and has industry research experience at Meta. His open-source work includes contributions to SuperPoint-related tooling, notably synthetic dataset generation and novel homography sampling used for robust local feature training. Comfortable across vision, video and text modalities, Rémi frequently bridges algorithmic research and practical dataset engineering to accelerate real-world spatial AI systems. He has a background in robotics and Bayesian optimization from early Inria work, reflecting a taste for probabilistic methods and resilient perception.
8 years of coding experience
2 years of employment as a software developer
Bac+2, French preparatory classes, Bac+2, French preparatory classes at Lycée du Parc, LyonMaster's degree, Informatique, Master's degree, Informatique at École PolytechniqueMaster in Computer Science, Computer Vision, Master in Computer Science, Computer Vision at ETH Zürich
français - french, anglais - english, allemand - german
Github Skills (5)
data-generation10
computer-vision10
opencv10
python10
numpy10
Programming languages (5)
Github contributions (5)
rpautrat/SuperPoint
Mar 2018 - Jan 2023
Efficient neural feature detector and descriptor
Role in this project:
Backend Developer & Data Scientist
Backend Developer & Data ScientistContributions:1 review, 166 commits, 27 PRs in 4 years 10 months
Contributions summary:Rémi was primarily involved in developing the synthetic dataset used in the repository. Their work focused on the creation of various geometrical shapes (lines, polygons, ellipses, etc.) and the implementation of noise and background generation functions using OpenCV (cv2). Further contributions included improvements in the generation of the checkerboard and the stripe patterns, along with vectorization of operations and a new approach to homography sampling.
deep-learningdescriptordetector
cvg/DeepLSD
Nov 2022 - Mar 2023
Implementation of the paper "DeepLSD: Line Segment Detection and Refinement with Deep Image Gradients"
Contributions:3 reviews, 19 commits, 5 PRs in 3 months
gradientsrefinementdeep-learningsegmentline-segment
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.